Aggressive behavior is a secondary symptom associated with dementia that is among the most difficult for nursing home staff to manage. The preliminary investigation proposed here within will serve as a feasibility and pilot study for a multiple site clinical trial that will address three long-range goals. These goals are: 1) to determine if implementing individualized care plans for the specific personal care activity of bathing prevents and reduces the occurrence of aggressive behavior in cognitively-impaired elderly nursing home residents known to be aggressive; 2) to evaluate if utilizing a geropsychiatric clinical nurse specialist (CNS) as a consultant and resource person to nursing assistants (NAs) during bathing situations is effective in enhancing their satisfaction with caregiving to aggressive residents and their implementation of individualized bathing care plans; and 3) to determine the efficiency of the model. Specific aims of the pilot study are: a) to document the precipitants and characteristics of verbal and physical aggressive behavior directed at staff during and after assistance with bathing; b) to document the process used by the CNS to develop individualized bathing care plans collaboratively with NAs for aggressive residents; c) to determine if the process used by the CNS increases NAs' satisfaction with caregiving, enhances the quality of their relationship with aggressive residents, and promotes their adoption of individualized bathing care plans; d) to determine if implementation of the plans decreases the frequency of verbal and physical aggressive behavior during and after bathing; and e) to develop and test the methods to be used in evaluating the efficiency of the model. A pretest-post test control group design will be used in which 20 residents and their NA caregivers will be randomly assigned to experimental or control groups at each of two sites (N=40). Quantitative data on resident, staff and cost outcomes will be collected pre and post-intervention and analyzed using repeated measures ANOVA. Qualitative data on the process and product of the intervention will be collected and analyzed using inductive methods and content analysis.